iNaturalist record replication speed and accuracy

What you describe is the sort of scenario predicted when the whole idea of importing iNat records was suggested.  'Research Grade' may be the only quality control available but as you demonstrate it is pretty meaningless.  When looking for species occurrence in an area I just ignore iNat records - either it is so common that they aren't needed or if it is uncommon then I don't trust them.

Agree that the speed of sharing record to iRecord from iNaturalist is too fast, as it is too easy for an inexperienced person on iNaturalist to offer an ID and the orginal recorder to accept it, making it research grade before any experienced person in the community has even had a chance to look at it. There needs to be more of a delay, in order for the community identification to have a better chance of correcting obvious errors.

Just trying out using iNat and agree that it is far too easy to get very dubious (totally wrong) records transferred too fast onto iRecord, especially of the many taxa that are not easy to identify from a single photo. Overall though it is better than I'd anticipated and it does make upload very easy. Frustrating to not be able to edit your own records though.

I have been using iNaturalist for a while and only recently became aware that many of these are now in iRecord. Confidential sites of my records in iNaturalist are assigned a random location in a 0.2 x 0.2 degree cell. However these have been simply carried over into iRecord as precise locations!  Also none of "my records" belong to me anymore because I have not imported them and there does not apear to be a way to filter on recorder. Not a very satisfactory situation.

@apout

Are you using iNaturalistUK or just plain iNaturalist? My understanding is that using the former means you agree to share all data with them, whereas the latter should not allow sharing obscured locations. In fact, I didn't think iRecord imported records with obscured locations at all from iNaturalist (as opposed to iNaturalistUK) - unless that has changed relatively recently. (NB: you can change your Network Affiliation in your iNat account settings).

If you are using iNaturalist (as opposed to iNaturalistUK) and you think your account settings somehow aren't being respected properly, you should report that to the iNat admins, as it would appear to be a breach of your privacy.

More generally, I think importing should be strictly opt-in, so that only users who explictly choose to join iNaturalistUK will have their records imported into iRecord.

The BRC are importing records from iNaturalist on a regular basis. Any changes made to a record in iNaturalist eg updating the observation licence to CC0, correcting a location or species ID will update the record in iRecord until it is verified. Once verified the observation record cannot be over ridden. 

The BRC are importing records from iNaturalist using the open API (See more here https://api.inaturalist.org/v1/docs/#/). This means that the records being imported are as available to view by anyone on the iNaturalist / iNaturalistUK website. So if users have obscured their location / made private these wont currently be imported into iRecord. Read more on data flow https://nbn.org.uk/inaturalistuk/inaturalistuk-and-its-place-in-biological-recording-data-flow/

Although the BRC and NBN Trust as leads for iNaturalistUK do have access to 'enhanced' data from observers who have affiliated to iNaturalistUK eg exact locations for sensitive species these are currently not being imported into iRecord. 

If I have already logged my sightings via iNaturalist, should I still log them onto iRecord as well, or would I be merely duplicating the information?

Records from iNaturalist that become research grade will be imported into iRecord, so there is no need to also log them on iRecord. There might be occasions where you would want to add records to iRecord directly, for example if you notice that there isn't anyone on iNaturalist giving IDs for a particular group so they never reach Research Grade and therefore never get transferred.

There are sometimes particular issues with records from iNaturalist, for example if you have a username rather than your actual name (e.g. "plantspotter" instead of "Andy T....") then some recording schemes might not accept the record, so it might be worth using the filters on iRecord to just check a few of the records you have submitted to iNaturalist by selecting the species, location and date just to check that they have imported correctly and if there is a verifier for that group then check it the record has been verified.

There's more to it than just Research Grade. The taxonomies used by iNaturalist and iRecord aren't fully compatible - anything that can't currently be found in the UK Species Inventory won't be imported into iRecord (and this is not a rare occurrence). For a complete list of the importing requirements, see here.

However, much more importantly, even if your records are imported, the verifiers are under no obligation to review them. And in fact, many verifiers have publicly stated that they will disregard any records coming from iNaturalist, regardless of their apparent quality. So if you really care about getting all of your records reviewed and accepted in iRecord, you should not rely on the automatic imports from iNaturalist.

The system is sufficiently robust to handle a certain amount of duplication, so it's not worth worrying about it too much. Personally, I care far more about maximising the chances that good use is made of my records. If data-sharing was much more reliable and feedback flowed in both directions, I would only ever use iNaturalist. Failing that, it would be really nice to opt out of automatic imports on a per-user basis, so the issue with potential duplication could be completely avoided. But I highly doubt iRecord will obtain the funding to implement any of that within my lifetime, so my best option is to continue uploading to both sites. The iRecord verifiers can easily filter out the iNaturalist imports if they want to avoid any duplicated effort.